Mechanisms of pancreatic cancer initiation and progression from normal human pancreatic tissue For the past thirty years, the survival rate for many cancers has improved, but survival for pancreatic ductal adenocarcinoma (PDAC) has not, due to lack of both early detection methods and effective treatments. Genetically modified mouse models have been widely used to study PDAC pathogenesis. However, mouse cancer models may overlook some profound differences between human and mouse cells. To understand the mechanisms of human PDAC tumorigenesis, we have established a novel system to isolate and genetically manipulate primary acinar cells and ductal cells from normal human pancreatic tissue. This unique system provides us with the opportunity to study the earliest changes in acinar or ductal cells during human PDAC initiation. Pancreatitis-induced acinar to ductal metaplasia (ADM) is considered to be the earliest change during PDAC development. Using our system, we recapitulated acinar to ductal metaplasia (ADM) in vitro at the single-cell level, and identified that the TGF? signaling pathway induces ADM in human acinar cells, partially through SMAD4-mediated pathway. During ADM, cells gained new characteristic properties, including transient proliferative capacity. However, the underlying mechanisms facilitating these changes remain unclear. We will use our system to further investigate the ADM process. KRAS is the earliest and most frequently mutated oncogene in pancreatic cancer. However, the mechanism by which oncogenic KRAS initiates neoplasia is not well understood, and it is generally accepted that additional genetic/epigenetic alterations are required to cooperate with KRAS mutation to initiate PDAC development. We found that mutant KRAS can extend the proliferation of AD cells (acinar cells that have undergone ADM). We hypothesize that ADM-associated changes cooperate with oncogenic KRAS to initiate pancreatic cancer. Although the SMAD4 pathway is required for ADM, this gene is frequently mutated in the late-stage PDAC samples, suggesting that it may have distinct functions under different cellular contexts. Interestingly, mouse PDAC models with SMAD4 deletion developed intraductal papillary-mucinous neoplasm (IPMN) rather than pancreatic intraepithelial neoplasia (PanIN). Whether this result is due to differences between mice and humans or due to the different cell lineages from which the tumors developed remains to be discovered. The findings from the proposed investigations will reveal human-specific aspects of PDAC tumorigenesis. The outcomes of these studies will not only provide an opportunity to discover prognostic markers but also give us insight into PDAC tumorigenesis that may lead to development of new methods for cancer treatment.